Evidence-based medicine


"Evidence" that a substance or procedure improved another person's health isn't evidence it will help yours.

Just as "evidence" of the effect of a substance given to a test animal is not evidence it will have the same effect on another test animal or on a human.

We are all unique and what works for one will not work for all.


True evidence-based medicine is medicine that you have used before and that worked for you.

That is evidence.

True evidence only relates to yourself in matters as personal as health. Population studies cannot tell you how this will be. Humans, after all, are not standardized robots, but unique living beings, no two the same.

How this argument would be disputed

People will dispute this by saying certain products have the same effect on everyone.  An example may be a poison that will kill anyone with a certain dosage level.

The reality is: unless every single human who ever lived was used as a test subject we cannot say a substance will effect everyone the same.  And even if all people could be tested, it would not mean someone born tomorrow would be the same.  Changes in human physiology do occur over time. This is called microevolution (evolution within a species).

Second dispute

This negation will be followed up by the idea that, true, we can never be sure of the effect of a substance on a person, but we can be confident because if it affects a high percentage of test subjects in a certain way, it is likely to affect a "random" other person in the same (or similar) way.

Here we are talking about statistics: the notion that what happened to a group of people means something about the likelihood of it occurring to you.

It should be realised this kind of idea is only real in a statistical universe.  It is not meaningful to an actual individual.

For the individual the odds of getting a certain cancer, for example, cannot be calculated.  We just don't know what will happen to an individual person.

The reality is, as far as disease is concerned, the odds of a specific person getting a certain thing cannot be calculated.

We can only truly speak of populations when we discuss statistics, and even then we can only speak of the population we have studied.

To do otherwise would involve making many assumptions which just aren't realistic to make (although statisticians do it themselves all the time out of necessity. We, non statisticians have no such necessity to make such false assumptions of conformity).

Such assumptions include: the sample of people which was tested is fundamentally the same (whatever that means) as the population as a whole.

This is impossible to prove and is pure conjecture used to justify statistical techniques.

Once a person contracts an illness we can say the odds of them receiving the illness as 100%, or P=1.  Until then, we just do not know, due to the uniqueness of the individual.

How this lack of certainty is dealt with

The lack of certainty about how a person will react to a "medicinal" substance is dealt with by letting a patient try a substance, check how it effects them and then move onto another substance if the first one doesn't have the preferred effect.

Eventually either (1) something is found to help the patient, (2) the patient heals himself or herself irrespective of medications, or, (3) the patient dies.

Comments